MLIP: Efficient Multi-Perspective Language-Image Pretraining with Exhaustive Data Utilization

TL;DR

MLIP enhances language-image pretraining by utilizing multi-perspective supervision through frequency and spatial domain analysis, improving data efficiency and reducing computational costs compared to traditional CLIP models.

Contribution

The paper introduces MLIP, a novel pretraining approach that leverages frequency transforms and token merging to provide richer supervision and efficiency improvements over CLIP.

Findings

MLIP achieves better performance on multimodal tasks.

It reduces computational costs by token merging.

Extensive experiments validate the effectiveness of MLIP.

Abstract

Contrastive Language-Image Pretraining (CLIP) has achieved remarkable success, leading to rapid advancements in multimodal studies. However, CLIP faces a notable challenge in terms of inefficient data utilization. It relies on a single contrastive supervision for each image-text pair during representation learning, disregarding a substantial amount of valuable information that could offer richer supervision. Additionally, the retention of non-informative tokens leads to increased computational demands and time costs, particularly in CLIP's ViT image encoder. To address these issues, we propose Multi-Perspective Language-Image Pretraining (MLIP). In MLIP, we leverage the frequency transform's sensitivity to both high and low-frequency variations, which complements the spatial domain's sensitivity limited to low-frequency variations only. By incorporating frequency transforms and token-level alignment, we expand CILP's single supervision into multi-domain and multi-level supervision, enabling a more thorough exploration of informative image features. Additionally, we introduce a token merging method guided by comprehensive semantics from the frequency and spatial domains. This allows us to merge tokens to multi-granularity tokens with a controllable compression rate to accelerate CLIP. Extensive experiments validate the effectiveness of our design.